Mathematical Sciences: Rank Tests for Multivariate Applications with Missing Data

点击翻译按钮获取中文摘要

The purpose of this project is to study a new class of rank statistics for analyzing nonserial dependent data with missing observations. The dependencies are assumed to come from multivariate data or repeated measures data. Nonparametric methods which do not involve the assumption of normally distributed data are to be developed. The testing and estimation procedures proposed will be based on asymptotic distributions of ranks of dependent data. They incorporate intercomponent and interblock rankings and, as a result, are applicable to a much wider class of problems with nonserial dependent data. They will also have improved power properties.